Created attachment 141011 [details] output of journalctl -xb Quite often (at this moment, I have even problems to finish this bug report, but normally it was just around five times daily) the screen gets black. All programs run (e.g., YouTube in Firefox is still playing), but I haven't found a way how to get screen reinitialized (Ctrl-Alt-F2 and similar do nothing). I observe this phenomenon for the last month or a little bit more, and it seems to be changing with upcoming upgrades of some packages. I would guess kernel, but I am not sure. Currently the situation is pretty bad (openSUSE Tumbleweed kernel 4.18.0-rc7-1.g68c9705-default), and it is somehow erratic. Once this happens, the computer tends to lock up all the time. Next day, computer is happy again until the first freeze happens again.
Created attachment 141012 [details] output of dmesg
Can you narrow down what package caused the regression? If so, can you bisect? There haven't really been many changes to the radeon kernel driver in the last few years.
(In reply to Alex Deucher from comment #2) > Can you narrow down what package caused the regression? All changes during two months on rather alive distribution? Sorry. > If so, can you bisect? Hardly, I am not sure I have 100% reliable reproducer. > There haven't really been many changes to the radeon kernel driver > in the last few years. Would some dumps from the graphic card or something help? How to obtain them?
(In reply to Matej Cepl from comment #3) > (In reply to Alex Deucher from comment #2) > > Can you narrow down what package caused the regression? > > All changes during two months on rather alive distribution? Sorry. > > > If so, can you bisect? > > Hardly, I am not sure I have 100% reliable reproducer. > > > There haven't really been many changes to the radeon kernel driver > > in the last few years. > > Would some dumps from the graphic card or something help? How to obtain them? Not really. Nothing looks out of the ordinary on your log. As for the symptoms, when you say black screen, do you mean the screen blanks like flickers or the signal to the monitor goes away or like someone has colored the desktop black? Please check that the display cable is not loose where it connects to the monitor and the graphics card.
(In reply to Alex Deucher from comment #4) > As for the > symptoms, when you say black screen, do you mean the screen blanks like > flickers or the signal to the monitor goes away or like someone has colored > the desktop black? Please check that the display cable is not loose where > it connects to the monitor and the graphics card. Like if somebody coloured the screen black. No flickering, and monitor doesn't complain about the loss of signal. And all programs run without an interruption (e.g., Firefox plays YouTube music without problems).
Any messages in the kernel log when the screen goes black?
Please attach the corresponding Xorg log file. My best guess so far would be this is bug 105381.
(In reply to Michel Dänzer from comment #7) > Please attach the corresponding Xorg log file. > > My best guess so far would be this is bug 105381. Do we have any Xorg.0.log on openSUSE? I thought we are journal only (thus log in the attachment 141011 [details] ).
(In reply to Matej Cepl from comment #8) > Do we have any Xorg.0.log on openSUSE? I thought we are journal only (thus > log in the attachment 141011 [details] ). I don't see the Xorg logging in there. Check in ~/.local/share/xorg/ .
Created attachment 141113 [details] ~/.local/share/xorg/Xorg.0.log I cannot reproduce right now so I am afraid these are not useful.
Created attachment 141114 [details] ~/.local/share/xorg/Xorg.0.log.old
Useful enough to confirm it's probably bug 105381, or maybe a known regression caused by the fix for that. Either way, assuming it's fixed with current upstream xf86-video-ati Git master. If it still happens with a snapshot of that which reports version 18.0.99 in the Xorg log, please reopen. *** This bug has been marked as a duplicate of bug 105381 ***
It turns out, the hardware might be faulty, so perhaps this is a PEBKAC in the end.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.